Investigate if pre-consensus time estimate is useful

A pre-consensus event has an estimated time for when it will reach consensus. This was originally added in an effort to make an educated guess what will the consensus time of an event be, so that we would handle pre-consensus events and consensus events in a similar order.

ATM, there is no estimate being done, so the functionality acts as a placeholder.

We need to evaluate the usefulness of this functionality and either implement good estimates or remove it completely.

Start by tracing how pre-consensus events calculate their estimated consensus time and how that estimate affects ordering relative to consensus events. Compare the current placeholder behavior with the intended ordering goal. Done means either a justified, working estimation approach is implemented or the unused functionality is removed completely.

Written by the indexing model from the issue text.